class: center, middle, inverse, title-slide # <span style='font-size: 70px'>Measuring and Predicting Player Fatigue in the NBA </style> ### <span style='font-size: 35px'>Austin Stephen, Matthew Yep and Grace Fain </style> ### <span style='font-size: 35px'> 7/13/2021 </style> --- ## Why fatigue? - Optimized Performance - Injury Prevention ??? 1. Add source that found as players get more tired performance worsens in basketball. Add source that talks about injury links to fatigue Add a source that --- ## The Plan 1. Measure player fatigue 2. Build a model that can be used to predict the impact of fatigue on a given game or multi-game sequence. ??? Identify a proxy for fatigue and link it to in game performance. This model can used by teams to examine their roster and match-ups and gain insights into how differnt factors are influencing their teams performance --- ## Measuring player fatigue - Cannot use direct measures - Identify a proxy ??? To be more concrete on what we mean by fatigue: "the decline in objective performance measures derived from the capacity of the nervous system and contractile properties of muscles over time" In a perfect world we would directly measure an NBA players internal homeostasis either with hematological or with sensors that measure heart rate. Howoever, players cannot wear heart rate monitors or tracking equipment in games and we certainly don't have data about in/post-game blood samples so we need a proxy to measure how fatigued a player was by in game events. --- ## Proxy for Fatigue - In game distance - Between game distance ??? We are examining two major sources for player fatigue. 1. The distance and speed at which they travel in game. We started by looking to validate this as proxy for player exertion in a given game. 2. The characteristics of their travel between games including days of rest, distance traveled and time zone changes. --- ## In-Game Distance Data <img src="presentation_files/figure-html/unnamed-chunk-1-1.png" width="100%" height="100%" />